arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2512.06253cs.CR
收藏

通过乘积测度的集中分析研究噪声扰动的隐私损失

Privacy Loss of Noise Perturbation via Concentration Analysis of A Product Measure

Texas Tech University(德州理工大学);Hangzhou Dianzi University(杭州电子科技大学)

Shuainan Liu, Tianxi Ji, Zhongshuo Fang, Lu Wei, Pan Li

AI总结:

本文提出基于乘积测度的噪声生成方案,通过几何视角研究噪声扰动的隐私损失,证明(ε,δ)-DP并提升高维噪声结果的效用。

摘要:

噪声扰动是实现查询或函数f(·)∈R^M在敏感数据集x上评估结果的(ε,δ)-差分隐私(DP)保证最基础的方法之一。在此方法中,校准的噪声n∈R^M用于掩盖差分向量f(x)-f(x'),其中x'称为邻近数据集。通过研究隐私损失随机变量(PLRV)的尾部概率界来获得DP保证,PLRV定义为两个分布之间的Radon-Nikodym导数。当n服从多变量高斯分布时,PLRV被表征为特定的单变量高斯分布。在本文中,我们提出了一种新的方案来生成n,利用扰动噪声通常球形对称(即分布围绕原点旋转不变)的事实。新的噪声生成方案使我们能够从几何角度研究隐私损失,并用乘积测度W×U表达所得到的PLRV;测度W与一个控制n大小的半径随机变量相关,而测度U涉及一个控制n与差分f(x)-f(x')之间角度的方向随机变量。我们推导了乘积测度的闭合形式矩界,以证明(ε,δ)-DP。在相同的(ε,δ)-DP保证下,我们的机制在高维情况下产生的噪声期望幅度比经典高斯噪声更小,从而显著提高了噪声结果f(x)+n的效用。为了验证这一点,我们考虑了高维空间中的凸和非凸经验风险最小化(ERM)问题,并应用所提出的乘积噪声来实现隐私。

英文摘要:

Noise perturbation is one of the most fundamental approaches for achieving $(ε,δ)$-differential privacy (DP) guarantees when releasing the result of a query or function $f(\cdot)\in\mathbb{R}^M$ evaluated on a sensitive dataset $\mathbf{x}$. In this approach, calibrated noise $\mathbf{n}\in\mathbb{R}^M$ is used to obscure the difference vector $f(\mathbf{x})-f(\mathbf{x}')$, where $\mathbf{x}'$ is known as a neighboring dataset. A DP guarantee is obtained by studying the tail probability bound of a privacy loss random variable (PLRV), defined as the Radon-Nikodym derivative between two distributions. When $\mathbf{n}$ follows a multivariate Gaussian distribution, the PLRV is characterized as a specific univariate Gaussian. In this paper, we propose a novel scheme to generate $\mathbf{n}$ by leveraging the fact that the perturbation noise is typically spherically symmetric (i.e., the distribution is rotationally invariant around the origin). The new noise generation scheme allows us to investigate the privacy loss from a geometric perspective and express the resulting PLRV using a product measure, $W\times U$; measure $W$ is related to a radius random variable controlling the magnitude of $\mathbf{n}$, while measure $U$ involves a directional random variable governing the angle between $\mathbf{n}$ and the difference $f(\mathbf{x})-f(\mathbf{x}')$. We derive a closed-form moment bound on the product measure to prove $(ε,δ)$-DP. Under the same $(ε,δ)$-DP guarantee, our mechanism yields a smaller expected noise magnitude than the classic Gaussian noise in high dimensions, thereby significantly improving the utility of the noisy result $f(\mathbf{x})+\mathbf{n}$. To validate this, we consider convex and non-convex empirical risk minimization (ERM) problems in high dimensional space and apply the proposed product noise to achieve privacy.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑